Week 6

Seaborn

import seaborn as sns
import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")
tips.head(5)

	total_bill	tip	sex	smoker	day	time	size
0	16.99	1.01	Female	No	Sun	Dinner	2
1	10.34	1.66	Male	No	Sun	Dinner	3
2	21.01	3.50	Male	No	Sun	Dinner	3
3	23.68	3.31	Male	No	Sun	Dinner	2
4	24.59	3.61	Female	No	Sun	Dinner	4

plt.scatter(tips["total_bill"], tips["tip"])

<matplotlib.collections.PathCollection at 0x7f9b1c5c4910>

sns.relplot(data=tips,
            x="total_bill",
            y="tip"
           )

<seaborn.axisgrid.FacetGrid at 0x7f9b1c5fa0d0>

sns.set_theme(style="darkgrid",
              palette="colorblind",
              font_scale=1.3 
             )

sns.relplot(data=tips,
            x="total_bill",
            y="tip"
           )

<seaborn.axisgrid.FacetGrid at 0x7f9b1c725fa0>

sns.relplot(data=tips,
            x="total_bill",
            y="tip", 
            height=6, 
            aspect=2
           )

<seaborn.axisgrid.FacetGrid at 0x7f9b1c8e2490>

sns.relplot(data=tips,
            x="total_bill",
            y="tip", 
            hue="day",
            height=6, 
            aspect=2
           )

<seaborn.axisgrid.FacetGrid at 0x7f9b1c6310a0>

sns.relplot(data=tips,
            x="total_bill",
            y="tip", 
            hue="day",
            size="size",
            sizes=(5, 400),
            height=6, 
            aspect=2
           )

<seaborn.axisgrid.FacetGrid at 0x7f9b1cfe9d30>

sns.relplot(data=tips,
            x="total_bill",
            y="tip", 
            hue="day",
            size="size",
            col="time",
            sizes=(5, 400),
            height=6, 
            aspect=1
           )

<seaborn.axisgrid.FacetGrid at 0x7f9b1d6e73a0>

sns.relplot(data=tips,
            x="total_bill",
            y="tip", 
            hue="day",
            size="size",
            col="time",
            row="sex",
            sizes=(5, 400),
            height=4, 
            aspect=1
           )

<seaborn.axisgrid.FacetGrid at 0x7f9b1db51370>

Figure-level and axes-level functions

g = sns.jointplot(data=tips, x="total_bill", y="tip", height=7)
g.set_axis_labels("Total bill", "Tip")
g.fig.suptitle("Restaurant visits", y=1.01)

Text(0.5, 1.01, 'Restaurant visits')

plt.figure(figsize=(10, 7))

plt.subplot(1, 2, 1)
sns.scatterplot(data=tips, 
                x="total_bill",
                y="tip",
                hue="day"
               )
plt.xlabel("Total bill")
plt.ylabel("Tip")
plt.plot([0, 50], [1, 10], 'r-', lw=5)

plt.subplot(1, 2, 2)
plt.plot([1, 2, 3, 4], [3, 0, 2, 1], 'bo-')

[<matplotlib.lines.Line2D at 0x7f9b1e54d820>]

Plotly Express

%pip install plotly

Requirement already satisfied: plotly in /Users/bb/opt/anaconda3/lib/python3.8/site-packages (4.14.1)
Requirement already satisfied: six in /Users/bb/opt/anaconda3/lib/python3.8/site-packages (from plotly) (1.15.0)
Requirement already satisfied: retrying>=1.3.3 in /Users/bb/opt/anaconda3/lib/python3.8/site-packages (from plotly) (1.3.3)
Note: you may need to restart the kernel to use updated packages.

import plotly.express as px

fig = px.scatter(data_frame=tips,
                 x="total_bill",
                 y="tip",
                 title="Restaurant data",
                 width=800,
                 height=400
                )
fig.show()

fig = px.scatter(data_frame=tips,
                 x="total_bill",
                 y="tip",
                 color="day",
                 title="Restaurant data",
                 width=1000,
                 height=500
                )
fig.show()

fig = px.scatter(data_frame=tips,
                 x="total_bill",
                 y="tip",
                 color="day",
                 size="size",
                 title="Restaurant data",
                 width=1000,
                 height=500
                )
fig.show()

fig = px.scatter(data_frame=tips,
                 x="total_bill",
                 y="tip",
                 color="day",
                 size="size",
                 facet_col="time",
                 title="Restaurant data",
                 width=1000,
                 height=500
                )
fig.show()

fig = px.scatter(data_frame=tips,
                 x="total_bill",
                 y="tip",
                 color="day",
                 size="size",
                 facet_col="time",
                 facet_row="sex",
                 title="Restaurant data",
                 width=1000,
                 height=500
                )
fig.show()

fig.write_html("my_plot.html")

Plotly

Line plots

gapminder = px.data.gapminder()
gapminder.sample(5)

	country	continent	year	lifeExp	pop	gdpPercap	iso_alpha	iso_num
861	Kuwait	Asia	1997	76.15600	1765345	40300.619960	KWT	414
511	Ethiopia	Africa	1987	46.68400	42999530	573.741314	ETH	231
611	Guatemala	Americas	2007	70.25900	12572928	5186.050003	GTM	320
1507	Taiwan	Asia	1987	73.40000	19757799	11054.561750	TWN	158
289	China	Asia	1957	50.54896	637408000	575.987001	CHN	156

ac = gapminder[gapminder["country"].str[0] == "A" ]
ac.sample(10)

	country	continent	year	lifeExp	pop	gdpPercap	iso_alpha	iso_num
62	Australia	Oceania	1962	70.930	10794968	12217.226860	AUS	36
30	Algeria	Africa	1982	61.368	20033753	5745.160213	DZA	12
19	Albania	Europe	1987	72.000	3075321	3738.932735	ALB	8
48	Argentina	Americas	1952	62.485	17876956	5911.315053	ARG	32
17	Albania	Europe	1977	68.930	2509048	3533.003910	ALB	8
67	Australia	Oceania	1987	76.320	16257249	21888.889030	AUS	36
27	Algeria	Africa	1967	51.407	12760499	3246.991771	DZA	12
9	Afghanistan	Asia	1997	41.763	22227415	635.341351	AFG	4
24	Algeria	Africa	1952	43.077	9279525	2449.008185	DZA	12
37	Angola	Africa	1957	31.999	4561361	3827.940465	AGO	24

fig = px.line(ac,
              x="year", 
              y="gdpPercap",
              color="country",
              hover_name="country",
              labels={"year": "Year", "gdpPercap": "GDP per capita", "country": "Country name"}
             )
fig.show()

Bar plot

fig = px.bar(tips,
             x="day",
             y="total_bill",
             color="sex",
             barmode="group",
             category_orders = {"day": ["Thur", "Fri", "Sat", "Sun"]}
            )
fig.show()

Strip plot

fig = px.strip(tips,
               x="day",
               y="tip",
               color="sex",
               category_orders = {"day": ["Thur", "Fri", "Sat", "Sun"]}
              )
fig.show()

Histogram

fig = px.histogram(tips, 
                   x='total_bill',
                   nbins=30,
                   histnorm="percent"
                  )

fig.update_layout({"bargap": 0.02})
fig.show()

Some other plots

fig = px.scatter_3d(tips,
                    x = "total_bill",
                    y="tip",
                    z="size",
                    color="day"
                   )
fig.show()

Choropleth maps

url = "https://raw.githubusercontent.com/plotly/datasets/master/2011_us_ag_exports.csv"

import pandas as pd

df = pd.read_csv(url)
df.head(5)

	code	state	category	total exports	beef	pork	poultry	dairy	fruits fresh	fruits proc	total fruits	veggies fresh	veggies proc	total veggies	corn	wheat	cotton
0	AL	Alabama	state	1390.63	34.4	10.6	481.0	4.06	8.0	17.1	25.11	5.5	8.9	14.33	34.9	70.0	317.61
1	AK	Alaska	state	13.31	0.2	0.1	0.0	0.19	0.0	0.0	0.00	0.6	1.0	1.56	0.0	0.0	0.00
2	AZ	Arizona	state	1463.17	71.3	17.9	0.0	105.48	19.3	41.0	60.27	147.5	239.4	386.91	7.3	48.7	423.95
3	AR	Arkansas	state	3586.02	53.2	29.4	562.9	3.53	2.2	4.7	6.88	4.4	7.1	11.45	69.5	114.5	665.44
4	CA	California	state	16472.88	228.7	11.1	225.4	929.95	2791.8	5944.6	8736.40	803.2	1303.5	2106.79	34.6	249.3	1064.95

df["cotton_exports"] = df["cotton"] > 0
df.head(5)

	code	state	category	total exports	beef	pork	poultry	dairy	fruits fresh	fruits proc	total fruits	veggies fresh	veggies proc	total veggies	corn	wheat	cotton	cotton_exports
0	AL	Alabama	state	1390.63	34.4	10.6	481.0	4.06	8.0	17.1	25.11	5.5	8.9	14.33	34.9	70.0	317.61	True
1	AK	Alaska	state	13.31	0.2	0.1	0.0	0.19	0.0	0.0	0.00	0.6	1.0	1.56	0.0	0.0	0.00	False
2	AZ	Arizona	state	1463.17	71.3	17.9	0.0	105.48	19.3	41.0	60.27	147.5	239.4	386.91	7.3	48.7	423.95	True
3	AR	Arkansas	state	3586.02	53.2	29.4	562.9	3.53	2.2	4.7	6.88	4.4	7.1	11.45	69.5	114.5	665.44	True
4	CA	California	state	16472.88	228.7	11.1	225.4	929.95	2791.8	5944.6	8736.40	803.2	1303.5	2106.79	34.6	249.3	1064.95	True

fig = px.choropleth(df,
                   locationmode="USA-states",
                   locations="code",
                   scope="usa",
                   color="cotton_exports",
                   color_discrete_sequence = ["red", "lightgray"],
                   title="States exporting cotton",
                   hover_name = "state", 
                   hover_data = {"cotton_exports": False, "code": False}
                  )
fig.update_layout(showlegend=False)
fig.show()

import plotly.express as px

df["veggies_perc"] = (df["total veggies"]/df["total exports"])*100
df.head(3)

	code	state	category	total exports	beef	pork	poultry	dairy	fruits fresh	fruits proc	total fruits	veggies fresh	veggies proc	total veggies	corn	wheat	cotton	veggies_perc
0	AL	Alabama	state	1390.63	34.4	10.6	481.0	4.06	8.0	17.1	25.11	5.5	8.9	14.33	34.9	70.0	317.61	1.030468
1	AK	Alaska	state	13.31	0.2	0.1	0.0	0.19	0.0	0.0	0.00	0.6	1.0	1.56	0.0	0.0	0.00	11.720511
2	AZ	Arizona	state	1463.17	71.3	17.9	0.0	105.48	19.3	41.0	60.27	147.5	239.4	386.91	7.3	48.7	423.95	26.443270

fig = px.choropleth(df, 
                    locationmode="USA-states",
                    scope="usa",
                    locations="code",
                    color="veggies_perc",
                    title="Precentage of veggie exports",
                    color_continuous_scale = "tempo",
                    hover_name="state",
                    labels={"veggies_perc": "Veggie %"},
                    hover_data = {"code": False}
                   )

fig.update_traces(marker_line_color="white")
fig.show()

dir(px.colors.sequential)[:10]

['Aggrnyl',
 'Aggrnyl_r',
 'Agsunset',
 'Agsunset_r',
 'Blackbody',
 'Blackbody_r',
 'Bluered',
 'Bluered_r',
 'Blues',
 'Blues_r']

Project: Baby names

from zipfile import ZipFile
import requests
import pandas as pd

# download names.zip
r = requests.get("https://www.ssa.gov/oact/babynames/names.zip")
with open("names.zip", 'wb') as foo:
    foo.write(r.content)

# read zipped csv file
with ZipFile("names.zip", 'r') as z:
    fnames = (f for f in z.namelist() if f.endswith(".txt"))
    file_dict = {}
    for f in fnames:
        with z.open(f) as foo:
            file_dict[int(f[3:-4])] = pd.read_csv(foo, names=["name", "sex", "count"])

# concatenate dataframes
df = pd.concat(file_dict.values(), keys=file_dict.keys())
df.reset_index(level=0, inplace=True)
df.rename({"level_0": "year"}, axis=1, inplace=True)

df

	year	name	sex	count
0	1880	Mary	F	7065
1	1880	Anna	F	2604
2	1880	Emma	F	2003
3	1880	Elizabeth	F	1939
4	1880	Minnie	F	1746
...	...	...	...	...
31949	2019	Zyheem	M	5
31950	2019	Zykel	M	5
31951	2019	Zyking	M	5
31952	2019	Zyn	M	5
31953	2019	Zyran	M	5

1989401 rows × 4 columns

n_babies = df.groupby(by="year")["count"].sum()
n_babies

year
1880     201484
1881     192696
1882     221533
1883     216944
1884     243461
         ...   
2015    3694784
2016    3661373
2017    3567022
2018    3503806
2019    3445321
Name: count, Length: 140, dtype: int64

px.line(x=n_babies.index, y=n_babies)

n_babies2 = df.groupby(by=["year", "sex"])["count"].sum()
n_babies2

year  sex
1880  F        90994
      M       110490
1881  F        91953
      M       100743
1882  F       107847
              ...   
2017  M      1845472
2018  F      1694640
      M      1809166
2019  F      1665373
      M      1779948
Name: count, Length: 280, dtype: int64

n_babies2 = n_babies2.reset_index()
n_babies2

	year	sex	count
0	1880	F	90994
1	1880	M	110490
2	1881	F	91953
3	1881	M	100743
4	1882	F	107847
...	...	...	...
275	2017	M	1845472
276	2018	F	1694640
277	2018	M	1809166
278	2019	F	1665373
279	2019	M	1779948

280 rows × 3 columns

px.line(n_babies2,
        x="year",
        y="count",
        color="sex"
       )

pd.pivot_table(n_babies2, index="year", columns="sex", values="count")

sex	F	M
year
1880	90994	110490
1881	91953	100743
1882	107847	113686
1883	112319	104625
1884	129019	114442
...	...	...
2015	1781725	1913059
2016	1767902	1893471
2017	1721550	1845472
2018	1694640	1809166
2019	1665373	1779948

140 rows × 2 columns

John

john = df[df["name"] == "John"]
john

	year	name	sex	count
228	1880	John	F	46
942	1880	John	M	9655
312	1881	John	F	26
938	1881	John	M	8769
277	1882	John	F	40
...	...	...	...	...
18426	2017	John	M	9499
7919	2018	John	F	14
18125	2018	John	M	9163
8290	2019	John	F	13
17932	2019	John	M	8779

280 rows × 4 columns

john = john.drop("name", axis=1)
john

	year	sex	count
228	1880	F	46
942	1880	M	9655
312	1881	F	26
938	1881	M	8769
277	1882	F	40
...	...	...	...
18426	2017	M	9499
7919	2018	F	14
18125	2018	M	9163
8290	2019	F	13
17932	2019	M	8779

280 rows × 3 columns

px.line(john, x="year", y="count", color="sex")

fig = px.line(john, x="year", y="count", color="sex", facet_row="sex")
fig.update_yaxes(matches=None)
fig.show()

n_babies2

	year	sex	count
0	1880	F	90994
1	1880	M	110490
2	1881	F	91953
3	1881	M	100743
4	1882	F	107847
...	...	...	...
275	2017	M	1845472
276	2018	F	1694640
277	2018	M	1809166
278	2019	F	1665373
279	2019	M	1779948

280 rows × 3 columns

john_tot = john.groupby(by="year")["count"].sum()
john_tot

year
1880     9701
1881     8795
1882     9597
1883     8934
1884     9428
        ...  
2015    10407
2016    10044
2017     9509
2018     9177
2019     8792
Name: count, Length: 140, dtype: int64

john_ratio = john_tot/n_babies
john_ratio

year
1880    0.048148
1881    0.045642
1882    0.043321
1883    0.041181
1884    0.038725
          ...   
2015    0.002817
2016    0.002743
2017    0.002666
2018    0.002619
2019    0.002552
Name: count, Length: 140, dtype: float64

px.line(x = john_ratio.index, y=john_ratio)

Most popular names

df

	year	name	sex	count
0	1880	Mary	F	7065
1	1880	Anna	F	2604
2	1880	Emma	F	2003
3	1880	Elizabeth	F	1939
4	1880	Minnie	F	1746
...	...	...	...	...
31949	2019	Zyheem	M	5
31950	2019	Zykel	M	5
31951	2019	Zyking	M	5
31952	2019	Zyn	M	5
31953	2019	Zyran	M	5

1989401 rows × 4 columns

def top_name(grp):
    return grp.sort_values(by="count", ascending=False).head(1)

most_popular = df.groupby(by=["year", "sex"]).apply(top_name)
most_popular

			year	name	sex	count
year	sex
1880	F	0	1880	Mary	F	7065
	M	942	1880	John	M	9655
1881	F	0	1881	Mary	F	6919
	M	938	1881	John	M	8769
1882	F	0	1882	Mary	F	8148
...	...	...	...	...	...	...
2017	M	18400	2017	Liam	M	18821
2018	F	0	2018	Emma	F	18757
	M	18100	2018	Liam	M	19915
2019	F	0	2019	Olivia	F	18451
	M	17905	2019	Liam	M	20502

280 rows × 4 columns

most_popular = most_popular.reset_index(level=2, drop=True)
most_popular

		year	name	sex	count
year	sex
1880	F	1880	Mary	F	7065
	M	1880	John	M	9655
1881	F	1881	Mary	F	6919
	M	1881	John	M	8769
1882	F	1882	Mary	F	8148
...	...	...	...	...	...
2017	M	2017	Liam	M	18821
2018	F	2018	Emma	F	18757
	M	2018	Liam	M	19915
2019	F	2019	Olivia	F	18451
	M	2019	Liam	M	20502

280 rows × 4 columns

most_popular = most_popular.drop(["year", "sex"], axis=1)

most_popular

		name	count
year	sex
1880	F	Mary	7065
	M	John	9655
1881	F	Mary	6919
	M	John	8769
1882	F	Mary	8148
...	...	...	...
2017	M	Liam	18821
2018	F	Emma	18757
	M	Liam	19915
2019	F	Olivia	18451
	M	Liam	20502

280 rows × 2 columns

most_popular["name"].unstack().T

year	1880	1881	1882	1883	1884	1885	1886	1887	1888	1889	...	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
sex
F	Mary	Mary	Mary	Mary	Mary	Mary	Mary	Mary	Mary	Mary	...	Isabella	Sophia	Sophia	Sophia	Emma	Emma	Emma	Emma	Emma	Olivia
M	John	John	John	John	John	John	John	John	John	John	...	Jacob	Jacob	Jacob	Noah	Noah	Noah	Noah	Liam	Liam	Liam

2 rows × 140 columns

presidents = pd.DataFrame([{"name": "Obama", "start": "2007", "end": "2017"},
                           {"name": "Trump", "start": "2017", "end": "2021"},
                          ])
presidents

	name	start	end
0	Obama	2007	2017
1	Trump	2017	2021

px.timeline(presidents,
            x_start="start",
            x_end="end",
            y="name"
           )

most_popular = most_popular.reset_index()
most_popular

	year	sex	name	count
0	1880	F	Mary	7065
1	1880	M	John	9655
2	1881	F	Mary	6919
3	1881	M	John	8769
4	1882	F	Mary	8148
...	...	...	...	...
275	2017	M	Liam	18821
276	2018	F	Emma	18757
277	2018	M	Liam	19915
278	2019	F	Olivia	18451
279	2019	M	Liam	20502

280 rows × 4 columns

most_popular["start_year"] = (most_popular["year"]).astype(str)
most_popular["end_year"] = (most_popular["year"] + 1).astype(str)
most_popular

	year	sex	name	count	start_year	end_year
0	1880	F	Mary	7065	1880	1881
1	1880	M	John	9655	1880	1881
2	1881	F	Mary	6919	1881	1882
3	1881	M	John	8769	1881	1882
4	1882	F	Mary	8148	1882	1883
...	...	...	...	...	...	...
275	2017	M	Liam	18821	2017	2018
276	2018	F	Emma	18757	2018	2019
277	2018	M	Liam	19915	2018	2019
278	2019	F	Olivia	18451	2019	2020
279	2019	M	Liam	20502	2019	2020

280 rows × 6 columns

fig = px.timeline(most_popular,
            x_start="start_year",
            x_end="end_year",
            y="name",
            color="sex",
            facet_row = "sex"
           )
fig.update_yaxes(matches=None)
fig.show()